AI Roadmap

A plan for using AI safely in your business from an IT firm that has been supporting businesses for 20+ years.

We help you understand your AI risks, possibilities, and what's worth doing. You get a roadmap: a prioritized plan to put AI to work where it counts. For Calgary and Alberta businesses with 15 to 100 employees.

Fixed-fee engagement, scoped to the size of your business.

Where most businesses are with AI

My team is using AI and I don't know what data is going into it.

Staff use personal ChatGPT or Claude accounts to draft emails and summarize documents, often with client or financial information in them. Microsoft Copilot, and any AI tool connected to your Microsoft 365, can see every file the user can see, so loose file permissions become searchable.

Nobody here has the time or background to lead this.

Deciding where AI fits across the business is a new job with no clear owner. IT already has a full plate, and the managers who know the business don't have the technical background.

We use five AI tools and none of them is part of how we work.

Tools get adopted one at a time by whoever found them. Results vary from person to person and day to day, and nothing connects to the systems the business runs on.

AI keeps showing up on our renewals and I can't tell what it's worth.

Microsoft, your accounting software, and your phone system are all adding AI features and charging for them, often at renewal. Add a few ChatGPT or Claude subscriptions on company cards, and you're paying for overlapping AI with no view of what gets used.

In a 2026 Goldman Sachs survey of 1,256 small business owners, 76% were using AI, but only 14% had it fully built into their core operations. (source)

How we decide what fits

Some of the work businesses want AI for is better handled by automation. Tasks that need the same result every time, such as moving data between systems, creating records from a form, or generating standard documents, run better on rules-based automation. It costs less to run and produces the same output every time. AI is the right tool when a person would normally have to read something and decide. Examples include sorting incoming requests, pulling details out of quotes and invoices, and answering staff questions from your own policies. The strongest results usually combine both: AI handles the judgement step, and automation does the rest.

Once we know what each task needs, we look at how to get it, starting with the least expensive option. The same three options apply to automation and to AI.

	Use what you have	Buy	Build
Automation	Turn on and set up features already in your software, and show staff how to use them	A product or add-on built for the job, for a monthly fee	Automations that connect your apps, up to a purpose-built app with the automation inside
AI	Set up the AI platform your team uses, such as Claude, ChatGPT, or Copilot, with shared instructions, company context, and training	An AI tool built for one job, such as meeting notes or call answering	An AI agent connected to your apps and data, up to a custom app with AI where it needs judgement and fixed rules where results must be consistent

Use what you have

Automation: Turn on and set up features already in your software, and show staff how to use them
AI: Set up the AI platform your team uses, such as Claude, ChatGPT, or Copilot, with shared instructions, company context, and training

Buy

Automation: A product or add-on built for the job, for a monthly fee
AI: An AI tool built for one job, such as meeting notes or call answering

Build

Automation: Automations that connect your apps, up to a purpose-built app with the automation inside
AI: An AI agent connected to your apps and data, up to a custom app with AI where it needs judgement and fixed rules where results must be consistent

Every option involves setup work. What changes is what you pay for: nothing new, a subscription, or something built for you. Each opportunity is scored on value to the business, effort to put in place, and whether it will hold up in daily use.

1. Your workflows
2. What each task needs: automation, AI, or both
3. How to get it: use what you have, buy, or build
4. Score: value, effort, reliability
5. Roadmap

What you get

Safe AI setup

Business AI accounts with sign-in through your Microsoft accounts, access controls, and a written AI usage policy for staff, with the AI governance settings that enforce it. Included in every engagement.

Workflow map

How work moves through your business today, with the time each role spends on it.

AI use review

Where staff use AI today, what information goes into it, and what needs to change.

Opportunity register

Every opportunity we find, tagged use, buy, or build, and scored on value, effort, and reliability.

Prioritized roadmap

What to do first, second, and later, with the approach and estimated cost for the first project.

IT budget benchmark

Your IT spend per user, compared with businesses of similar size.

Why an IT firm

AI is software that connects to your data and acts with whatever access it's given, like a new employee with a company laptop. We manage it with the same rules we apply to every user and device: access limited to what the job requires, sign-in through company accounts, and monitoring for changes.

We know what sticks
Years of supporting business software and the people who use it show us which tools get used and which get worked around.
Inside your IT controls
Access, security settings, and change management are part of the design from the start.
Supported after launch
We support what we put in place, bought or built.

Engagement options

The engagement comes in three levels, and each includes everything in the level below it. Every engagement is quoted as a fixed fee, scoped to the size of your business.

Map

Workflow interviews
Time and cost mapping
Review of your systems and data
AI use review
Safe AI setup
Prioritized roadmap
IT budget benchmark

Recommended

Equip

Everything in Map, plus:

Up to four hands-on sessions with your team
Working prototypes of real workflows, built on your AI platform
Prototypes your team keeps and uses

Partner

Everything in Equip, plus:

30 days of advisory support after the roadmap
Vendor evaluation for anything you buy
Configuration guidance for tools you already own
Support for your first project as it starts

How it works

1. Call: 30 minutes to understand your business and confirm the engagement fits.
2. Setup: Safe AI setup and the interview schedule.
3. Interviews and mapping: We talk with the people doing the work and map where their time goes.
4. Sort and score: Each task is matched to automation or AI, given a use, buy, or build option, and ranked.
5. Roadmap: We walk you through the plan and the first project.

Kickoff to roadmap takes 4 to 6 weeks.

Questions

Personal AI accounts, free or paid, are controlled by the individual, not the business. Staff can upload files or connect them to email and shared drives, and the business has no record of what was shared. The engagement sets up business accounts with settings that control what AI can connect to and which features are on. Anything else stays off until it's been evaluated. A written policy then sets the rules: business accounts on approved platforms only, never personal ones.

If staff use AI tools at all, yes. Even on a company platform with central controls, those controls can't cover everything, so a policy sets the baseline everyone follows. It covers what's allowed, what information stays out of AI tools, and who to ask. AI can draft a generic policy in seconds. But a rule like "only use approved AI tools" means nothing until someone decides what's approved and why. The engagement does that work, so the policy fits how your business runs.

It depends on how your team works and how your files are permissioned. Copilot can see every file the user can see, so permissions need attention before it's turned on. The roadmap shows whether Copilot is worth the licence cost for each role, or whether another tool or a feature you already own does the job.

Where they want to be, yes. We work alongside internal IT staff or your current provider, and the roadmap includes a section written for them covering access, security settings, and systems.

That's a common result. Many opportunities are better handled by automation or by features in software you already own. The roadmap says so and estimates costs accordingly.

A readiness assessment typically checks whether your systems and data are ready to adopt AI tools. The AI Roadmap includes that check as a starting point. From there, it maps your workflows and the time each role spends on them, then recommends whether to use what you have, buy, or build. The result is a costed plan, which may or may not include a new AI product.

Start with a 30-minute call

We'll talk through how your team uses AI today and where you want to take it, then recommend which level fits. If the engagement isn't the right fit, we'll say so.